1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
// Copyright 2017 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.
//! A version of the Naive datalog analysis using Datafrog.
use datafrog::{Iteration, Relation, RelationLeaper};
use std::time::Instant;
use crate::facts::FactTypes;
use crate::output::{Context, Output};
pub(super) fn compute<T: FactTypes>(
ctx: &Context<'_, T>,
result: &mut Output<T>,
) -> (
Relation<(T::Loan, T::Point)>,
Relation<(T::Origin, T::Origin, T::Point)>,
) {
let timer = Instant::now();
let (errors, subset_errors) = {
// Static inputs
let origin_live_on_entry_rel = &ctx.origin_live_on_entry;
let cfg_edge = &ctx.cfg_edge;
let loan_killed_at = &ctx.loan_killed_at;
let known_placeholder_subset = &ctx.known_placeholder_subset;
let placeholder_origin = &ctx.placeholder_origin;
// Create a new iteration context, ...
let mut iteration = Iteration::new();
// .. some variables, ..
let subset = iteration.variable::<(T::Origin, T::Origin, T::Point)>("subset");
let origin_contains_loan_on_entry =
iteration.variable::<(T::Origin, T::Loan, T::Point)>("origin_contains_loan_on_entry");
let loan_live_at = iteration.variable::<((T::Loan, T::Point), ())>("loan_live_at");
// `loan_invalidated_at` facts, stored ready for joins
let loan_invalidated_at = Relation::from_iter(
ctx.loan_invalidated_at
.iter()
.map(|&(loan, point)| ((loan, point), ())),
);
// different indices for `subset`.
let subset_o1p = iteration.variable_indistinct("subset_o1p");
let subset_o2p = iteration.variable_indistinct("subset_o2p");
// different index for `origin_contains_loan_on_entry`.
let origin_contains_loan_on_entry_op =
iteration.variable_indistinct("origin_contains_loan_on_entry_op");
// Unfortunately, we need `origin_live_on_entry` in both variable and relation forms:
// We need:
// - `origin_live_on_entry` as a Relation for the leapjoins in rules 3 & 6
// - `origin_live_on_entry` as a Variable for the join in rule 7
//
// The leapjoins use `origin_live_on_entry` as `(Origin, Point)` tuples, while the join uses
// it as a `((O, P), ())` tuple to filter the `((Origin, Point), Loan)` tuples from
// `origin_contains_loan_on_entry_op`.
//
// The regular join in rule 7 could be turned into a `filter_with` leaper but that would
// result in a leapjoin with no `extend_*` leapers: a leapjoin that is not well-formed.
// Doing the filtering via an `extend_with` leaper would be extremely inefficient.
//
// Until there's an API in datafrog to handle this use-case better, we do a slightly less
// inefficient thing of copying the whole static input into a Variable to use a regular
// join, even though the liveness information can be quite heavy (around 1M tuples
// on `clap`).
// This is the Naive variant so this is not a big problem, but needs an
// explanation.
let origin_live_on_entry_var =
iteration.variable::<((T::Origin, T::Point), ())>("origin_live_on_entry");
origin_live_on_entry_var.extend(
origin_live_on_entry_rel
.iter()
.map(|&(origin, point)| ((origin, point), ())),
);
// output relations: illegal accesses errors, and illegal subset relations errors
let errors = iteration.variable("errors");
let subset_errors = iteration.variable::<(T::Origin, T::Origin, T::Point)>("subset_errors");
// load initial facts:
// Rule 1: the initial subsets are the non-transitive `subset_base` static input.
//
// subset(Origin1, Origin2, Point) :-
// subset_base(Origin1, Origin2, Point).
subset.extend(ctx.subset_base.iter());
// Rule 4: the issuing origins are the ones initially containing loans.
//
// origin_contains_loan_on_entry(Origin, Loan, Point) :-
// loan_issued_at(Origin, Loan, Point).
origin_contains_loan_on_entry.extend(ctx.loan_issued_at.iter());
// .. and then start iterating rules!
while iteration.changed() {
// Cleanup step: remove symmetries
// - remove origins which are `subset`s of themselves
//
// FIXME: investigate whether is there a better way to do that without complicating
// the rules too much, because it would also require temporary variables and
// impact performance. Until then, the big reduction in tuples improves performance
// a lot, even if we're potentially adding a small number of tuples
// per round just to remove them in the next round.
subset
.recent
.borrow_mut()
.elements
.retain(|&(origin1, origin2, _)| origin1 != origin2);
// Remap fields to re-index by keys, to prepare the data needed by the rules below.
subset_o1p.from_map(&subset, |&(origin1, origin2, point)| {
((origin1, point), origin2)
});
subset_o2p.from_map(&subset, |&(origin1, origin2, point)| {
((origin2, point), origin1)
});
origin_contains_loan_on_entry_op
.from_map(&origin_contains_loan_on_entry, |&(origin, loan, point)| {
((origin, point), loan)
});
// Rule 1: done above, as part of the static input facts setup.
// Rule 2: compute the subset transitive closure, at a given point.
//
// subset(Origin1, Origin3, Point) :-
// subset(Origin1, Origin2, Point),
// subset(Origin2, Origin3, Point).
subset.from_join(
&subset_o2p,
&subset_o1p,
|&(_origin2, point), &origin1, &origin3| (origin1, origin3, point),
);
// Rule 3: propagate subsets along the CFG, according to liveness.
//
// subset(Origin1, Origin2, Point2) :-
// subset(Origin1, Origin2, Point1),
// cfg_edge(Point1, Point2),
// origin_live_on_entry(Origin1, Point2),
// origin_live_on_entry(Origin2, Point2).
subset.from_leapjoin(
&subset,
(
cfg_edge.extend_with(|&(_origin1, _origin2, point1)| point1),
origin_live_on_entry_rel.extend_with(|&(origin1, _origin2, _point1)| origin1),
origin_live_on_entry_rel.extend_with(|&(_origin1, origin2, _point1)| origin2),
),
|&(origin1, origin2, _point1), &point2| (origin1, origin2, point2),
);
// Rule 4: done above as part of the static input facts setup.
// Rule 5: propagate loans within origins, at a given point, according to subsets.
//
// origin_contains_loan_on_entry(Origin2, Loan, Point) :-
// origin_contains_loan_on_entry(Origin1, Loan, Point),
// subset(Origin1, Origin2, Point).
origin_contains_loan_on_entry.from_join(
&origin_contains_loan_on_entry_op,
&subset_o1p,
|&(_origin1, point), &loan, &origin2| (origin2, loan, point),
);
// Rule 6: propagate loans along the CFG, according to liveness.
//
// origin_contains_loan_on_entry(Origin, Loan, Point2) :-
// origin_contains_loan_on_entry(Origin, Loan, Point1),
// !loan_killed_at(Loan, Point1),
// cfg_edge(Point1, Point2),
// origin_live_on_entry(Origin, Point2).
origin_contains_loan_on_entry.from_leapjoin(
&origin_contains_loan_on_entry,
(
loan_killed_at.filter_anti(|&(_origin, loan, point1)| (loan, point1)),
cfg_edge.extend_with(|&(_origin, _loan, point1)| point1),
origin_live_on_entry_rel.extend_with(|&(origin, _loan, _point1)| origin),
),
|&(origin, loan, _point1), &point2| (origin, loan, point2),
);
// Rule 7: compute whether a loan is live at a given point, i.e. whether it is
// contained in a live origin at this point.
//
// loan_live_at(Loan, Point) :-
// origin_contains_loan_on_entry(Origin, Loan, Point),
// origin_live_on_entry(Origin, Point).
loan_live_at.from_join(
&origin_contains_loan_on_entry_op,
&origin_live_on_entry_var,
|&(_origin, point), &loan, _| ((loan, point), ()),
);
// Rule 8: compute illegal access errors, i.e. an invalidation of a live loan.
//
// Here again, this join acts as a pure filter and could be a more efficient leapjoin.
// However, similarly to the `origin_live_on_entry` example described above, the
// leapjoin with a single `filter_with` leaper would currently not be well-formed.
// We don't explictly need to materialize `loan_live_at` either, and that doesn't
// change the well-formedness situation, so we still materialize it (since that also
// helps in testing).
//
// errors(Loan, Point) :-
// loan_invalidated_at(Loan, Point),
// loan_live_at(Loan, Point).
errors.from_join(
&loan_live_at,
&loan_invalidated_at,
|&(loan, point), _, _| (loan, point),
);
// Rule 9: compute illegal subset relations errors, i.e. the undeclared subsets
// between two placeholder origins.
// Here as well, WF-ness prevents this join from being a filter-only leapjoin. It
// doesn't matter much, as `placeholder_origin` is single-value relation.
//
// subset_error(Origin1, Origin2, Point) :-
// subset(Origin1, Origin2, Point),
// placeholder_origin(Origin1),
// placeholder_origin(Origin2),
// !known_placeholder_subset(Origin1, Origin2).
subset_errors.from_leapjoin(
&subset,
(
placeholder_origin.extend_with(|&(origin1, _origin2, _point)| origin1),
placeholder_origin.extend_with(|&(_origin1, origin2, _point)| origin2),
known_placeholder_subset
.filter_anti(|&(origin1, origin2, _point)| (origin1, origin2)),
// remove symmetries:
datafrog::ValueFilter::from(|&(origin1, origin2, _point), _| {
origin1 != origin2
}),
),
|&(origin1, origin2, point), _| (origin1, origin2, point),
);
}
// Handle verbose output data
if result.dump_enabled {
let subset = subset.complete();
assert!(
subset
.iter()
.filter(|&(origin1, origin2, _)| origin1 == origin2)
.count()
== 0,
"unwanted subset symmetries"
);
for &(origin1, origin2, location) in subset.iter() {
result
.subset
.entry(location)
.or_default()
.entry(origin1)
.or_default()
.insert(origin2);
}
let origin_contains_loan_on_entry = origin_contains_loan_on_entry.complete();
for &(origin, loan, location) in origin_contains_loan_on_entry.iter() {
result
.origin_contains_loan_at
.entry(location)
.or_default()
.entry(origin)
.or_default()
.insert(loan);
}
let loan_live_at = loan_live_at.complete();
for &((loan, location), _) in loan_live_at.iter() {
result.loan_live_at.entry(location).or_default().push(loan);
}
}
(errors.complete(), subset_errors.complete())
};
info!(
"analysis done: {} `errors` tuples, {} `subset_errors` tuples, {:?}",
errors.len(),
subset_errors.len(),
timer.elapsed()
);
(errors, subset_errors)
}